Scott Adams reads Newsweek. Uh-oh.

If a little knowledge is a dangerous thing, the insignificant, minute information Adams has on evolution must be exceedingly risky—it’s like the atom bomb of ignorance. In this case, it’s not entirely his fault, though. He read the recent Newsweek cover story on evolution, which fed his biases and readily led him smack into the epicenter of his own blind spots, and kerblooiee, he exploded.

This is a case where the flaws in a popular science article neatly synergize with an evolution-denialist’s misconceptions to produce a perfect storm of stupidity.

I’m not really impressed with the Newsweek article. It gives a nice overview of some current productive lines of research, but it’s too infatuated with ‘revolution disease’. It’s not enough that the new evidence is expanding our knowledge and supporting evolutionary theory—it all has to have the extra ingredient of fake drama added. The story is full of “we used to think this, and gosh, now we now that that was all wrong!” nonsense which gets in the way of an accurate message.

The science of human evolution is undergoing its own revolution. Although we tend to see the march of species down through time as a single-file parade, with descendant succeeding ancestor in a neat line, the emerging science shows that the story of our species is far more complicated than Biblical literalists would have it–but also more complex than secular science suspected.


Ahem…”we”, kemosabe? That diagram to the right is a famous sketch from Darwin’s notebooks, drawn in 1837. This idea that our history is a linear series of begats is a biblical misconception, not something that evolutionary biologists have endorsed, and they have never favored that simplistic pattern, not even since The Origin was a glimmering in the mind of one man. Now way back in the 19th and early 20th century, there was a time when the number of known species in the hominid lineage was small enough that we could imagine a simple anagenetic succession—if you’ve only got two points, it’s easy to draw a line—but that was fairly quickly dispensed with as more and more diversity was unearthed. Why, even when I was a wee young bairn getting all my paleoanthropology from the Louis Leakey stories in National Geographic, the message was that there were all these different species coexisting simultaneously in ancient Africa.

And what’s with this “secular science” throw-away line? There’s no other kind.

As another for instance, after explaining the lovely work by Mark Stoneking on louse phylogeny, the reporter has to throw this in:

If you had asked paleoanthropologists a generation ago what lice DNA might reveal about how we became human, they would have laughed you out of the room.

Why, no. I guarantee you, they would have said, “That’s brilliant!” A generation ago they would have also rightly said that there are a lot of technical hurdles to overcome, but anthropologists are smart people who have experience with learning about human evolution by studying those little incidentals and cast-offs and artifacts of human life; supplementing our knowledge of humans with knowledge of human parasites would have been readily accepted. If you want to see a responsible explanation of the importance of studying parasite evolution, try reading Carl Zimmer’s take on a related issue; writers really do not have to drag out this lazy trope of pretending scientists mocked a good idea in order to make it exciting.

I groaned a little every time the reporter pulled this faux controversy crap — it deeply marred an otherwise interesting article on some really cool work. The message ought to have been one of refinement and increasing detail in understanding human evolution, not revolution. And it ought to have been more appreciative of the fact that working out the peculiarities of the history of one lineage is most definitely not the same as overturning theoretical principles that have been worked out for everything from bacteria to trees—there’s more than a little narrow parochialism in the article.

Unfortunately, the buggered sensationalism of the article’s presentation is all that the profoundly ignorant are going to take away from it. Take Scott Adams, for example; he read that article, and all he seems to have learned is the observation that scientific interpretations of hominid fossils have changed over the years. Forget new, deeper, broader information—science has CHAAAAAAAANGED! Change is bad!

I’ve been trying for years to reconcile my usually-excellent bullshit filter with the idea that evolution is considered a scientific fact. Why does a well-established scientific fact set off my usually-excellent bullshit filter like a five-alarm fire? It’s the fossil record that has been bugging me the most. It looks like bullshit. Smells like bullshit. Tastes like bullshit. Why isn’t it bullshit? All those scientists can’t be wrong.

If you are new to the Dilbert Blog, I remind you that I don’t believe in Intelligent Design or Creationism or invisible friends of any sort. I just think that evolution looks like a blend of science and bullshit, and have predicted for years that it would be revised in scientific terms in my lifetime. It’s a hunch – nothing more.

Yesterday I read this article in Newsweek about how DNA testing is being used to show that, well, fossils are bullshit.

No, the article does not say that, anywhere, or in any way. Fossils are still an essential part of explaining our history. DNA is also an essential part of the story. Putting the two together generates an even stronger collection of evidence for the evolution of our species.

I read his little whine, and I’m afraid that the lesson he should have taken from this is that his bullshit detector isn’t excellent—in fact, it’s badly broken.

The bottom line is that DNA tests (which do not set off my bullshit detector) have shown that you can’t really tell what set of bones begat other sets of bones just by looking at how they differed and how old they are. Apparently evolution is more complex than imagined, and there were lots of ape-people varieties wandering around at the same time. Some had modern features that they weren’t supposed to have. The so-called modern features apparently popped up and disappeared more than once, and in more than one species.

DNA tests do not set off his bullshit detector? Now I know it’s broken! I think we’ve got another fellow who has fallen under the spell of the CSI reality distortion field: DNA is not magic. Interpreting DNA is as difficult and requires as much specialized training, and is as subject to revision, as interpreting bones. What we have are multiple strains of evidence, each acting as a reality check on each other, each with their own strengths and weaknesses. The only reason he accepts “DNA tests” is that Scott Adams for once recognizes that he knows nothing at all about a technological tool; he only thinks he knows it all about fossils.

Evolution has always been more complex than Scott Adams imagined. That statement indicts Adams’ imagination, not evolutionary biology.

As for the examples he cites…it’s curious that most of those complexities he mentions are derived from the fossil evidence, not the molecular evidence. His bullshit detector seems to be remarkably inconsistent.

To be fair, there’s still plenty of evidence for evolution. It’s not going away anytime soon. But personally, I’m cautious about any theory that keeps the same conclusion regardless of how many times the evidence for it changes. There was a time when the seemingly straight line of fossil evidence was the primary foundation for the theory. Now it seems that that straight line was like Little Billy from Family Circus finding his way home from the playground.

The “seemingly straight line of fossil evidence” never existed and was never the “primary foundation for the theory”. See diagram above. Darwin. 1837. Branchy branchy branchy. Read Gould’s Full House(amzn/b&n/abe/pwll). 1997. Evolution is like a drunkard’s walk (hmmm, I never thought about that before, but has Little Billy been drinking?)

And there was a time when it seemed evolution was probably a fairly continuous and gradual process. Now it seems it happened in bursts, relatively speaking.

This has always been a subject for much discussion, and variation has been recognized. See George Gaylord Simpson. 1947. Tempo and Mode in Evolution(amzn/b&n/abe/pwll). Again, Adams mistakes his erroneous preconceptions for the actual scientific debate.

And there was a time when it seemed that mutations had to give some sort of survival advantage to endure, and now scientists believe that isn’t necessarily true.

No, we know that isn’t true, and we’ve known that for a good long while. Read Larry Moran’s essay on drift, for instance. Fisher came up with the math behind it in 1922; Wright named the phenomenon of drift in 1931. How old is Scott Adams, anyway?

It’s clear what Adams’ problems are. The bullshit detector he takes such pride in could function as a random number generator; he’s got this weird idea that revising interpretations on the basis of better evidence implies that entire disciplines of science are false; he imagines that his knowledge of biology is state-of-the-art; and at the same time, he’s abysmally ignorant. It’s a good thing that his profession requires him to act like a buffoon, because that combination of flaws guarantees that he’s very, very good at it.

Adams has put up a new post in which he claims to have made successful predictions. One is this:

1. DNA evidence shows that ape-human fossil records have been badly misinterpreted. (Nailed it.)

The man is unreachable by reason or evidence. And the rest of his ‘predictions’ are equally divorced from reality — he seems quite fond of “The Secret”, for instance.


  1. #1 David Marjanovi?
    March 17, 2007

    Trouble was, Sahelanthropus tchadensis (nicknamed Toumai, the local word for “child”) lived close to 7 million years ago. The genetic data, pointing to a human-chimp split at least 1 million years later

    With an error margin of how much? +- 2 million years?

    I’ve done molecular dating. If you don’t choose enough calibration points in the right time range and the right parts of the tree, the results are wrong.

    Worse yet, in this particular case I can’t see any potential calibration points. (Can anyone else?) This means that *shudder* a constant rate must have been taken from elsewhere. Firstly, evolution rates are not constant, so this approach gives a large error range; secondly, the rate that was amazingly popular for birds — used as late as 2006 — has turned out to have been calculated under wrong assumptions (the split between two goose species had happened much earlier than assumed). If that rate was used here, the date must have been underestimated.

    Of course, it’s not impossible that either convergence is at work or that the chimps have undergone a couple of reversals. But it’s not likely.

  2. #2 David Marjanovi?
    March 18, 2007

    It’s frequently calibrated on when orangutangs split from other great apes (for which there are likely fossils).

    Ah, good. Still just one point, and external, but clearly better than nothing.

    Why would anyone use a bird-derived rate on apes?

    Good question, but I’ve seen worse.

    The current models can accommodate rate differences between lineages and all sorts of other detail. Of course, this doesn’t necessarily make it any easier to choose the best models.

    Yes, I’ve done some molecular dating… and fortunately stumbled over a paper that showed that Modeltest uses a neighbor-joining guide tree… I digress.

    Then there was the interpretation of humans and chimps interbreeding over millions of years.

    The evidence for this is quite good, judging from the paper.

    In short, as yourself, do some physical anthropologists define the human ancestor as an ape? And do apes have an ancestor that we can place under the definition of monkey if we so desire?

    Yes, except you can’t define something as an ape — you can only apply the definition of “ape” (if there is one!) to it and see if it fits.

    (Now, there’s nothing particularly wrong with elitism. I admit it: I’m an elitist. I love my elite so much I think everybody should belong to it.)

    Great quote!

    BTW, Augustine is a troll. He has posted this screed on several blogs, always in completely unrelated threads.

  3. #3 David Marjanovi?
    March 19, 2007

    Like using the artiodactyl/cetacean split as a calibration point for hominids?

    OUCH! Yes, this kind of thing. (Never mind that the artiodactyls are paraphyletic.)

    The worst I can remember off the top of my head is using the “primate-rodent divergence” 110 Ma ago as a calibration point for all manner of mammals. That has happened several times and was based on the Nature paper by Kumar & Hedges (1998) where the rodents were polyphyletic, mice and rats being the first placental branches to diverge due to painfully obvious long-branch attraction, and 110 Ma being the average age of these two nodes based on the single calibration point of the theropsid-sauropsid divergence 310 Ma ago (a date that is at least a bit too young and very poorly constrained). In short, the use of a fictitious calibration point.

    Only he dresses it up by saying that humans made humans with some BS called retrocasuality, where I guess future people created past people, thereby ensuring that the future people would exist to create the past people…


    That IS humor. He IS just being funny.